Molecular Geometry
Molecular geometry defines the three-dimensional spatial arrangement of atomic nuclei at a local energy minimum on the Born-Oppenheimer potential energy surface, governed by steric numbers, point-group symmetry, and electronic structure.
Molecular geometry (or molecular structure) is the three-dimensional spatial arrangement of the constituent atomic nuclei within a covalently bound molecule or polyatomic ion. Within the framework of quantum mechanics, a discrete molecular geometry emerges only after invoking the Born–Oppenheimer approximation (formulated by Max Born and J. Robert Oppenheimer in 1927), which decouples fast electronic motion from slow nuclear motion because atomic nuclei are 1,836 to >105 times more massive than electrons. Solving the electronic Schrödinger equation at fixed nuclear coordinates generates a multidimensional potential energy surface (PES), E(R), on which stable chemical species correspond to local minima where the net force on every nucleus vanishes (∇*E* = 0) and all eigenvalues of the mass-weighted Hessian matrix are positive. While valence shell electron pair repulsion (VSEPR) theory and orbital hybridization predict idealized coordination polyhedra from valence electron counts, the physical molecular geometry reports exclusively the positions of the atomic nuclei, distinguishing it from the broader electron-domain geometry that also encompasses non-bonding lone pairs.
Historically, the concept that molecules possess definite three-dimensional architectures originated in 1874, when Jacobus Henricus van 't Hoff and Joseph Achille Le Bel independently proposed that the four valence bonds of a saturated carbon atom are directed toward the vertices of a regular tetrahedron (θ = 109.47°), explaining optical isomerism in tartaric acid and substituted methanes. In 1893, Alfred Werner extended three-dimensional stereochemistry to inorganic coordination complexes, demonstrating octahedral and square-planar geometries around transition metal centers. During the 1920s and 1930s, the advent of X-ray crystallography, gas-phase electron diffraction, and high-resolution microwave rotational spectroscopy enabled direct experimental measurement of internuclear distances to sub-picometer precision.
Molecular geometry governs the physical, spectroscopic, and biochemical properties of matter. The vector sum of individual bond dipoles across a molecule's three-dimensional framework dictates whether the species possesses a permanent electric dipole moment (μ), directly controlling dielectric constants, boiling points, and intermolecular solvation forces. Furthermore, in biological systems, the sub-angstrom complementarity between small-molecule substrates and enzyme active sites or receptor binding pockets depends on precise bond lengths, bond angles, and torsional conformations.
---
Internal Coordinates and Degrees of Freedom
To specify the position of N atomic nuclei in three-dimensional Cartesian space, a total of 3*N* coordinates (xi, yi, zi for i = 1, ..., N) are required. However, three of these degrees of freedom describe rigid translation of the center of mass along the x, y, and z axes, and three (or two for a collinear arrangement of nuclei) describe rigid overall rotation of the molecule in space. Consequently, the intrinsic molecular geometry is completely defined by the remaining internal vibrational degrees of freedom:
ninternal = 3*N* − 6 (for non-linear molecules)
ninternal = 3*N* − 5 (for linear molecules)
For example, the non-linear triatomic water molecule (H2O, N = 3) possesses 3(3) − 6 = 3 internal coordinates, whereas the linear carbon dioxide molecule (CO2, N = 3) possesses 3(3) − 5 = 4 internal modes (the symmetric stretch, asymmetric stretch, and a doubly degenerate pair of orthogonal bending modes). In computational chemistry and structural crystallography, these internal coordinates are parameterized as a Z-matrix comprising three geometric primitives:
- Bond Lengths (rij): The equilibrium Euclidean distance between two covalently bonded nuclei i and j, typically expressed in angstroms (1 Å = 10−10 m = 100 pm) or picometers (pm). Bond length decreases as bond order increases and as the principal quantum number n of the valence shell decreases: for carbon–carbon bonds, r(C–C) ≈ 154 pm (sp3), r(C=C) ≈ 134 pm (sp2), and r(C≡C) ≈ 120 pm (sp).
- Bond Angles (θijk): The interior angle formed by two adjacent bonds sharing a common central vertex atom j, defined by the dot product cos θijk = (rji · rjk) / (|*r*ji| |*r*jk|).
- Dihedral (Torsion) Angles (φijkl): For any chain of four consecutively bonded atoms i–j–k–l, the angle between the plane containing atoms (i, j, k) and the plane containing atoms (j, k, l), corresponding to rotation about the central j–k bond. While covalent bond stretching and angle bending require tens to hundreds of kilojoules per mole, torsional rotation about single σ bonds typically encounters low energy barriers (e.g., 12.1 kJ mol−1 for the staggered-to-eclipsed barrier in ethane, C2H6), permitting rapid conformational interconversion at ambient thermal energy (kBT ≈ 2.48 kJ mol−1 at 298 K).
---
Electron-Domain vs. Molecular Geometry Across Steric Numbers
Under the Gillespie–Nyholm AXmEn classification scheme, the steric number (SN = m + n) equals the sum of σ-bonded peripheral substituent atoms (X, count m) and valence non-bonding lone pairs (E, count n) around a central atom A. The SN electron domains adopt a polyhedral arrangement that minimizes Pauli exclusion and electrostatic repulsion, establishing the electron-domain geometry. Because diffraction and spectroscopic techniques locate atomic nuclei rather than diffuse electron clouds, the molecular geometry describes only the spatial figure formed by A and the m peripheral X nuclei.
| Steric Number (SN) | AXmEn Class | Electron-Domain Geometry | Molecular Geometry | Ideal / Typical Bond Angles (θ) | Schoenflies Point Group (Symmetrical X) | Representative Chemical Examples |
|---|---|---|---|---|---|---|
| 2 | AX2E0 | Linear | Linear | 180° | D∞h (or C∞v) | CO2, BeCl2, HCN, C2H2 |
| 3 | AX3E0 | Trigonal planar | Trigonal planar | 120° | D3h | BF3, SO3, NO3−, CO32− |
| 3 | AX2E1 | Trigonal planar | Bent (angular) | 115°–119.5° | C2v | SO2 (119.5°), O3 (116.8°), NO2− (115°) |
| 4 | AX4E0 | Tetrahedral | Tetrahedral | 109.47° | Td | CH4, CCl4, NH4+, SO42− |
| 4 | AX3E1 | Tetrahedral | Trigonal pyramidal | 93°–107.8° | C3v | NH3 (107.8°), PF3 (97.8°), H3O+ |
| 4 | AX2E2 | Tetrahedral | Bent (angular) | 92°–104.45° | C2v | H2O (104.45°), OF2 (103.1°), H2S (92.1°) |
| 5 | AX5E0 | Trigonal bipyramidal | Trigonal bipyramidal | 90° (ax), 120° (eq) | D3h | PF5, PCl5(g), Fe(CO)5 |
| 5 | AX4E1 | Trigonal bipyramidal | Seesaw (disphenoidal) | <90° (ax), <120° (eq) | C2v | SF4 (101.6° eq, 173.1° ax), TeCl4 |
| 5 | AX3E2 | Trigonal bipyramidal | T-shaped | <90° (87.5°) | C2v | ClF3, BrF3 |
| 5 | AX2E3 | Trigonal bipyramidal | Linear | 180° | D∞h | XeF2, I3−, IC12− |
| 6 | AX6E0 | Octahedral | Octahedral | 90°, 180° | Oh | SF6, Mo(CO)6, PF6− |
| 6 | AX5E1 | Octahedral | Square pyramidal | <90° (84.8°) | C4v | BrF5, IF5, XeOF4 |
| 6 | AX4E2 | Octahedral | Square planar | 90°, 180° | D4h | XeF4, ICl4− |
Lone-Pair Compression and Bent's Rule
In the tetrahedral (SN = 4) series, the idealized vertex angle derived from setting the dot product of two unit vectors from the centroid of a cube to alternate corners yields θ = arccos(−1/3) ≈ 109.4712° (109° 28'). Replacing bonding pairs with non-bonding lone pairs progressively compresses the remaining X–A–X bond angles: methane (CH4, AX4) exhibits θ = 109.47°, ammonia (NH3, AX3E1) contracts to 107.8°, and water (H2O, AX2E2) contracts further to 104.45°. Because a lone pair is attracted by only one atomic nucleus rather than shared between two, its electron density spreads out laterally close to the central core, exerting greater angular repulsion according to the sequence:
Lone Pair–Lone Pair (LP–LP) > Lone Pair–Bonding Pair (LP–BP) > Bonding Pair–Bonding Pair (BP–BP)
Simultaneously, Bent's rule (formulated by Henry A. Bent in 1961) provides the quantum-mechanical basis for these angular deviations: a central atom directs hybrid orbitals with higher s character toward more electropositive substituents (or into non-bonding lone pairs) to lower orbital energy, while directing hybrids enriched in higher-energy p character toward more electronegative ligands. Because the angle between two equivalent hybrid orbitals with fractional p-to-s ratio λ2 obeys cos θ = −1/λ2, concentrating s character in the lone pair of H2O leaves the two O–H bonding hybrids with elevated p character, narrowing the H–O–H angle toward the 90° limit of pure orthogonal p orbitals. In heavier congeners such as H2S (θ = 92.1°) and PH3 (θ = 93.5°), the large energy gap between the valence 3*s* and 3*p* subshells suppresses sp3 hybridization, leaving the lone pair predominantly in the spherical *n*s orbital while bonding occurs through nearly unhybridized orthogonal *n*p orbitals.
Equatorial Preference in Five- and Six-Coordinate Geometries
Unlike regular polyhedra (SN = 2, 3, 4, 6) in which all vertices are symmetry-equivalent, the trigonal bipyramid (SN = 5, D3h) possesses two chemically distinct coordination sites: two axial positions collinear with the principal C3 axis (each experiencing three close 90° interactions with the equatorial plane) and three equatorial positions in the horizontal σh plane (each experiencing only two 90° axial interactions and two wider 120° in-plane interactions). Because inter-domain repulsion decays sharply with angular separation, sterically bulky groups, π-donating oxo ligands, and non-bonding lone pairs invariably occupy the less crowded equatorial vertices in SN = 5 species, generating the seesaw (AX4E1, e.g., SF4), T-shaped (AX3E2, e.g., ClF3), and linear (AX2E3, e.g., XeF2) molecular geometries. Conversely, in octahedral SN = 6 systems, all six vertices are equivalent; introducing one lone pair gives a square-pyramidal AX5E1 geometry (BrF5), while a second lone pair locates trans (180°) to the first to minimize LP–LP 90° interactions, producing the planar AX4E2 geometry of XeF4.
---
Point-Group Symmetry, Dipole Moments, and Fluxionality
Schoenflies Point Groups and Molecular Polarity
Every static molecular geometry belongs to a mathematical Schoenflies point group defined by its complete set of symmetry operations—proper n-fold rotation axes (Cn), mirror planes (σv, σh, σd), inversion centers (i), and improper rotation-reflection axes (Sn)—that leave at least one point (the center of mass) invariant. Point-group symmetry imposes rigorous selection rules on molecular polarity. Each polar covalent bond between atoms of differing electronegativity contributes an individual bond dipole vector μi = qeff ri. The net molecular electric dipole moment is the vector resultant μnet = Σ μi + μLP.
By symmetry, any permanent electric dipole vector must lie simultaneously along every rotation axis and within every mirror plane of the molecule. Consequently, a molecule can possess a nonzero permanent dipole moment (μ ≠ 0) if and only if it belongs to one of the point groups C1, Cs, Cn, or Cnv (including C∞v). In centrosymmetric or higher-symmetry point groups—such as D∞h (CO2), D3h (BF3, PF5), Td (CF4), D4h (XeF4), and Oh (SF6)—the peripheral bond dipoles cancel identically by vector addition (μnet = 0 D), even when the individual A–X bonds are highly polar. Conversely, lowering the symmetry through lone-pair distortion (H2O in C2v, μ = 1.85 D; NH3 in C3v, μ = 1.47 D) or asymmetric ligand substitution (CHCl3 in C3v, μ = 1.04 D) prevents vector cancellation.
Stereochemical Non-Rigidity and Fluxionality
Although molecular geometry is defined at a potential energy minimum, molecules with low barrier heights between symmetry-equivalent minima exhibit stereochemical non-rigidity (fluxionality) on spectroscopic timescales:
- Pyramidal (Umbrella) Inversion: Trigonal pyramidal AX3E1 molecules such as ammonia (NH3) undergo rapid quantum-mechanical tunneling through a planar D3h transition state (where the lone pair transiently occupies an unhybridized 2*p*z orbital). In NH3, the inversion barrier is only 24.2 kJ mol−1, causing the molecule to invert at a microwave frequency of 23.786 GHz (the basis of Charles Townes's 1954 ammonia maser). By contrast, heavier pnictogen hydrides and phosphines (PH3, AsH3, P*R*1R2R3) possess barriers of 130–155 kJ mol−1, rendering chiral phosphines configurationally stable at room temperature.
- Berry Pseudorotation: Pentacoordinate D3h molecules such as phosphorus pentafluoride (PF5) and iron pentacarbonyl (Fe(CO)5) rapidly exchange their two axial and three equatorial ligands intramolecularly via a C4v square-pyramidal transition state (Berry pseudorotation, proposed by R. Stephen Berry in 1960). Because the activation barrier in PF5 is only ~15 kJ mol−1, 19F nuclear magnetic resonance (NMR) spectroscopy observes a single time-averaged fluorine doublet down to −100 °C, whereas rapid gas-phase electron diffraction (10−18 s interaction time) resolves the distinct rax (157.7 pm) and req (153.4 pm) bond lengths.
---
Experimental Determination of Molecular Geometry
Because real molecules undergo continuous zero-point and thermal vibrations even at 0 K, experimental bond lengths and angles depend slightly on the physical observable and vibrational averaging inherent to each analytical method:
- Microwave Rotational Spectroscopy (r0 and rs): Measures quantized rotational transitions (Δ*J* = ±1) of gas-phase polar molecules (μ ≠ 0). The rotational constant B = h / (8*π*2cI) yields the principal moments of inertia I = Σ mi ri2, where rotational kinetic energy follows Erot = J(J + 1)*ℏ*2 / (2*I*). Isotopic substitution (Kraitchman's equations) pinpoints nuclear coordinates to ±0.05 pm, allowing extrapolation to the theoretical bottom-of-the-well equilibrium bond length (re).
- Gas-Phase Electron Diffraction (rg): Scatters a monochromatic beam of relativistic electrons (40–60 keV) off an effusive molecular jet, yielding the thermally averaged internuclear distance rg for both polar and non-polar volatile compounds.
- X-Ray and Neutron Crystallography: Single-crystal X-ray crystallography diffracts off the total atomic electron density, locating heavy-atom centers accurately while systematically foreshortening X–H bond lengths (by ~10 pm) because the single valence electron of hydrogen is polarized into the covalent bond. Neutron diffraction, which scatters directly off atomic nuclei via the strong nuclear force, determines true proton positions and hydrogen-bond geometries with high accuracy.
---
Frequently Asked Questions
What is the difference between electron-domain geometry and molecular geometry?
Electron-domain geometry describes the spatial arrangement of all valence electron regions—both covalent bonding pairs and non-bonding lone pairs—around a central atom, whereas molecular geometry describes only the three-dimensional positions of the atomic nuclei. When a central atom has no lone pairs (such as CH4 or SF6), the two geometries are identical. When lone pairs are present (such as H2O, AX2E2), the electron-domain geometry is tetrahedral, but the molecular geometry of the nuclei is bent.
Why is the bond angle in water (104.45°) smaller than in ammonia (107.8°) and methane (109.47°)?
Methane, ammonia, and water all share a steric number of four (SN = 4), based on a tetrahedral distribution of four valence electron pairs. However, non-bonding lone pairs are held by only a single atomic nucleus rather than stretched between two nuclei, causing their electron clouds to occupy greater angular volume near the central core. Simultaneously, by Bent's rule, lone pairs concentrate higher 2*s* orbital character, leaving the remaining O–H or N–H bonding hybrids with elevated 2*p* character and narrower interbond angles.
Why do lone pairs occupy equatorial rather than axial positions in trigonal bipyramidal molecules?
In a trigonal bipyramidal (SN = 5) framework, the axial and equatorial vertices are geometrically non-equivalent. An axial domain sits at 90° to all three equatorial neighbors, whereas an equatorial domain sits at 90° to only the two axial neighbors and at 120° to the other two equatorial domains. Because Pauli exchange repulsion between electron pairs drops off steeply as the angle increases from 90° to 120°, placing large lone pairs in the equatorial plane minimizes short-range 90° interactions.
---
References
- Gillespie, R. J., & Nyholm, R. S. (1957). "Inorganic stereochemistry." Quarterly Reviews, Chemical Society, 11(4), 339–380. https://doi.org/10.1039/QR9571100339
- Bent, H. A. (1961). "An appraisal of valence-bond structures and hybridization in compounds of the first-row elements." Chemical Reviews, 61(3), 275–311. https://doi.org/10.1021/cr60211a005
- Berry, R. S. (1960). "Correlation of rates of intramolecular tunneling processes, with application to some Group V compounds." The Journal of Chemical Physics, 32(3), 933–938. https://doi.org/10.1063/1.1730820
- Gillespie, R. J. (2008). "Fifty years of the VSEPR model." Coordination Chemistry Reviews, 252(12–14), 1315–1327. https://doi.org/10.1016/j.ccr.2007.07.007